Organizers

Event Contact

Programme

Time Speaker Description
14.00 - 16.00 Mireille Hildebrandt Tutorial on the proposal for an AI Act

Background

All partners need to prepare for the tutorial, made easy by a small library of presentations that

- discuss the most important players, concepts, structure and obligations in the proposal

The presentations consist of slides with audio, explaining the text.

The library can be found at the internal service of the HAI network

During the session Hildebrandt will present a general introduction to the proposal, highlighting its architecture, links and deeplinks with the existing framework (product safety) and the upcoming framework (Digital Market Act, Digital Services Act, Data Governance Act). This introduction will form slide-set 0, which will be added to the library after the event.

TUTORIAL Library:

A series of Slide-sets with Audio:

    1. AIA General provisions, definitions and prohibitions
      Title I – art. 1-4, Annex I and Title II – art. 5
    2. AIA What are high-risk AI systems?
      Title III, chapter 1 – art. 6-7, Annex II and III
    3. AIA Risk management system for high risk systems
      Title III, chapter 2 – art. 9
    4. AIA Data and data governance of high risk systemsWhat information must be provided to whom and how?
      Title III, chapter 2 – art. 10
    5. AIA Transparency of high risk systems
      Title III, chapter 2 – art. 13
    6. AIA Human oversight of high risk systems
      Title III, chapter 2 – art. 14
    7. AIA Accuracy, robustness and cybersecurity of high risk systems
      Title III, chapter 2 – art. 15
    8. AIA Obligations of providers of high risk systems
      Title III, chapter 3 – art. 16-23
    9. AIA Obligations for users of high risk systems
      Title III, chapter 3 – art. 29
    10. AIA Notification, conformity assessment of high risk systems
      Title III, chapter 5 – art. 42-44 and 48-49
    11. AIA Transparency for medium risk systems
      Title IV – art. 52
    12. AIA Remaining issues

Organizers

Event Contact

Programme

Time Speaker Description
12.00 - 13.30 Mireille Hildebrandt

Background

This is an internal event, where all partners need to prepare themselves, based a small library of presentations that:

      1. introduce the tutorial and
      2. provide a small set of concepts (and legal norms) core to the GDPR. The presentations consist of slides with audio, explaining the text. The library can be found at the internal webserver of the HAI-NET

The library also provides access to the Textbook Law for Computer Scientists and Other Folk that contains relevant literature, notably in chapters 5 and 10.

TUTORIAL Library:

 The Open Access Textbook:

      • Law for Computer Scientists and Other Folk (OUP 2020, available in Open Access)
      • See dedicated sections below

A series of Slide-sets with Audio:

      • A 45-minute slide-set with audio: Main Introduction to the Tutorial
        Check out the introduction, the glossary and study chapter 5 of the Textbook
      • A set of eight short slide-sets with audio that introduce
        core GDPR requirements
        1. What is the legal status of developers, users and end-users?
          Controllers and processors (art. 4 GDPR)
          Section 5.5.2.3 and 5.5.2.4 of the Textbook
        2. On what ground can you process personal data?
          Legal basis (art. 6 GDPR)
          Section 5.2.5 of the Textbook
        3. What rules inform lawful development and use of data-driven AI?
          Principles (art. 5 GDPR)
          Section 5.2.6 of the Textbook
        4. What counts as valid consent?
          Consent (art. 7 GDPR)
          Section 5.2.7 of the Textbook
        5. What information must be provided to whom and how?
          Transparency (art. 12-15 GDPR)
          Section 5.4.1 GDPR of the Textbook (though at generic level)
        6. What kind of automated decisions are prohibited by default?
          Automated decisions (art. 22 GDPR)
          Section 3.3.3 of the Textbook (in relation to e.g. DLTs)
        7. How to embed legal norms in systems, architectures and applications?
          Data protection by design and default (DPbDD) (art. 25 GDPR)
          Section 5.2.9 and 10.3.3.2 of the Textbook
        8. When and how to assess the impact of data-driven AI applications?
          Data protection impact assessment (DPIA) (art. 35 GDPR)
          Section 5.2.10 and 10.3.3.1 of the Textbook

Organizers

Event Contact

Programme

Time Speaker Description
14:00 - 14:10 Roberto Trasarti SoBigData++ project: an ecosystem for Ethical Social Mining - This talk introduces SoBigData++ project with the aim of putting in context the participants presenting the main objectives of the project and the consortium of experts involved working on the vertical contextes: Societal Debates and Online Misinformation, Sustainable Cities for Citizens, Demography, Economics & Finance 2.0, Migration Studies, Sports Data Science, Social Impact of Artificial Intelligence and Explainable Machine Learning. Part of this presentation will be the description of an ethical approach to data science which is a pillar of the SoBigData++ project.
14:10 - 14:25 Valerio Grossi SoBigData RI Services - An overview of the SoBigData RI services will be shown including the Exploratories (Vertical research contexts), the resource catalogue, the training area and SoBigData Lab.
14:25 - 14:55 Giulio Rossetti Hands-on JupyterHub service and SoBigData Libraries - This first hands-on session focuses on the libraries and methods developed within the SoBigData consortium. Code examples and case studies will be introduced by leveraging a customized JupyterHub notebook service hosted by SoBigData. Using such a freely accessible coding environment, we will discuss a subset of the functionalities available to SoBigData users to design and run their experiments.
14:55 - 15:10 Massimiliano Assante Hands-on computational engine & technologies - In this second hands-on session, the tutorial will focus on the computational engine provided by SoBigData. Real examples will be presented in order to highlight the functionalities to deploy an algorithm and run it on the cloud.
15:10 - 15:25 Giovanni Comandè Legality Attentive data Science: it is needed and it is possible!
15:25 - 15:35 Francesca Pratesi FAIR: an E-learning module for GDPR compliance and ethical aspects
15:35 - 16:00 Beatrice Rapisarda (moderator) An open discussion to give more details on specific aspects according to the requests of the audience (not already addressed during the tutorial or presentations).

Objectives

The objectives of the tutorial are to show how SoBigData RI can support data scientists in doing cutting-edge science and experiments. In this perspective, our target audience also includes people interested in big data analytics, computational social science, digital humanities, city planners, wellbeing, migration, sport, health within the legal/ethical framework for responsible data science and artificial intelligence applications. With its tools and services, SoBigData RI promotes the possibilities that new generations of researchers have for executing large-scale experiments on the cloud making them accessible and transparent to a community. Moreover, specialized libraries developed in SoBigData++ project will be freely accessible in order to make cutting edge science in a cross-field environment.

Format: The tutorial will be 3 hours containing:

  • 1 hour of presentations describing the European project SoBigData++, the RI Services, and the Responsible Data Science principles and tools;
  • 45 minutes and half of practical use of the RI with real examples of analysis in a dedicated Virtual research environment;
  • 20 minutes for an open discussion with the attendees on the various aspects presented.